v1.2.9

• pip packages updated to barney 0.9.10
  -> better support for unstructured data, now according to (upcoming) ANARI 1.1 spec (including cell centric data)
• fixed missing anari.UINT8 type (needed for unstructured)
• fixed missing anari.FLOAT32_BOX1 type (volume value range)
• added sample/test case for unstructured data with vertex centric data (samples/unstructured-vertexCentric.py)
• added sample/test case for unstructured data with cell centric data (samples/unstructured-cellCentric.py)
• added sample/test case for setting value range (part of unstructured-cellCentric)

• fixed shape/array dimension ordering from xyz to zyx in all numpy shape() to anari array conversions
• updated all examples for fixed array ordering
• integrated new interactive viewer sample supplied by @jar091
• new frame::readGPU method that can read frame buffer into existing device buffer (for pycuda/pygl users)
• pip packages upated to barney 0.9.8

For pynari repo:

• added ability to map() frames, including ablity to map gpu side buffers if supported by device
• more samples, including non-trivial orb
• various updates/fixes to existing samples
• fixes to how data arrays can be created from numpy files
  For PIP Wheels:
• PIP wheels updated to latest barney 0.9.6